I asked an AI chatbot to tell me about myself. It was wrong in an uncanny way.

Like many people, I’ve been playing around with AI chat programs. I wasn’t sure really how I could get one to be useful. I’ve asked one to revise my CV, and it just moved some bullet points around and made it sound a bit more punchy. Fine, but not really a game changer. I asked them to write me a cover letters for jobs, and the results were both bombastic and obsequious — I am the best, most perfect candidate ever, for your esteemed and highly prestigious position — hardly a combination I’d like to convey.

Then I spoke to a friend who uses AI all the time for her tech job. Her advice was to “use it like an intern.” So I thought about what I would ask an intern — if I ever had one — to do. I’m planning to go to Spain, so I asked it to find me a price on business-class tickets to Barcelona. No dice. Apparently that’s not in its repertoire (and any other “intern” activities I could think of, like ordering lunch or sorting files, obviously weren’t either). 

Rather running out of ideas but still curious, I asked Claude — a chatbot developed by the startup Anthropic, in which Amazon has invested $4 billion — what it knew about me.

At first it said it hadn’t heard of me, which seemed odd; not that I’m famous, but I have an unusual enough name and my own website, so I thought I’d be findable. I then told it that I’ve written for the Wall Street Journal and the Washington Post. And with those crumbs of information, we were off to the races. Suddenly it generated a whole biography. 

It spat out this:

Katrina Gulliver is an American freelance journalist who has contributed reporting and articles to major publications like The Wall Street Journal and The Washington Post.

Some key points about her work:

– She has covered national security, cybersecurity, and technology topics for WSJ and WaPo among other outlets.

– Notable stories include reporting on the Anom encrypted phone sting operation against criminal organizations run by the FBI.

– She has written extensively about cybercrime, hacking, and the activities of hacker groups and cybercriminals.

– Her reporting has provided in-depth accounts of major cyberattacks, data breaches, and the tactics of nation-state hacking groups.

– Gulliver has expertise covering the dark web and illicit online marketplaces as well.

– Based in Washington D.C., she was previously a staff writer at The Hill newspaper covering technology policy issues.

– Her freelance work has also appeared in publications like Wired, Motherboard/Vice, and Bloomberg News.

So in summary, Katrina Gulliver is an accomplished cybersecurity and national security journalist with a background writing for respected newspapers and outlets in that space. Let me know if you need any other specific details regarding her reporting areas.

The only bit that’s true here is the word “freelance.” I’m a historian by training, and my journalism is mostly culture-related. While I’ve reviewed a few books in the cybercrime/surveillance area, I’ve never done any reporting in that area. 

Intriguingly (or disturbingly), it hovered closer to the truth on one point. I was writing a review of a book about the Anom sting operation; but it had not been published at this point, or mentioned anywhere online. I’d been told AIs could “hallucinate,” but this was oddly specific.

Yet somehow the AI built me a whole life. A staff writer position! (Never had one of those, at The Hill or anywhere else). Reading this description, I sound impressive. Like some kind of reporting ninja, the kind who goes on CNN or gives presentations on cybercrime at SXSW. Not the kind of person who needs to send obsequious cover letters.

I thought about it as almost a “Sliding Doors” version of myself. What I might have been in another life, the one in which I became an investigative reporter and not a historian? But AI imagined it for me. If I’d wanted to embellish my resume to sound more impressive, I never would have chosen that angle.

We’ve moved from working to present our most flattering selves online (the filtered selfie, photoshopping scars and messy hair), to even outsourcing that. AI can now generate a series of vacation pics where I’m in a bikini under a waterfall, and I haven’t been anywhere.

In the future, you don’t edit your life. AI edits you.

Katrina Gulliver is a historian and freelance writer. She writes the Substack “Notes from the Field.”